Last week the BBC released their list of the most streamed songs in the UK, split by year and, for the data geeks amongst us, it revealed some surprising patterns. Bohemian Rhapsody was the only one of the UK’s top 20 selling singles of all time to make the list and not one of the three songs that have spent the most weeks in top spot made the list at all.
No-one brought this difference into sharper focus than Canada’s favourite son, Bryan Adams. Whilst the Robin Hood ballad, 'Everything I Do (I Do It For You)', spent 16 weeks at the top of the charts in 1991, it has been usurped by Nirvana’s grunge anthem 'Smells Like Teen Spirit' in the Beeb’s list. The ‘Groover from Vancouver’ need not be too forlorn though, making the list with BBQ/ karaoke anthem ‘Summer of ‘69’ , a song which peaked at a lowly No. 88 when it was first released in 1985.
A quick analysis of the peak chart position of each of the songs on the list (I know I need to get out more!) shows that the older the song, the weaker the relationship between its chart position at the time and the likelihood to appear on the list. Obviously there is both a recency and an age bias at play here with streaming sites more likely to be used by younger music fans, listening to more recent music.
However, it's clear there is more to it than that. High selling songs central to the zeitgeist when released (e.g. 'Candle in the Wind' or 'Relax') don’t seem so relevant many years or decades later and understandably novelty tracks end up consigned to the same fate. Conversely, songs which build deeper memory structures (e.g. through movies) have a higher chance of standing the test of time as do those which are covered or even sampled, exposing the song to a fresh generation of ears (e.g. Ray Charles “I’ve Got a Woman”) .
However, it’s not just in the world of music where things can feel different with the passing of time. In his excellent book, Don’t Trust Your Gut, Seth Stephens-Davidowitz compares the factors which are proven to be important on dating sites vs. those which predict long term romantic happiness. Sadly, but perhaps unsurprisingly, it's physical factors such as height, race and classic ‘good looks’ that are all strong predictors of being ‘swiped right’ on dating sites. It’s not just about physical appearance with attributes such as income also playing a key role; indeed, a man could up his propensity to be contacted by nearly 9% simply by changing his income from $35k-$50k to $150-$200k.
Yet, just like music, these initial drivers of attraction are typically terrible predictors as to whether a relationship will last the test of time. Research by the University of Texas, quoted in the same book, shows that, rather than these superficial characteristics, it is actually deeper dimensions such as satisfaction with life, conscientiousness or a growth mindset which are better predictors of long term romantic happiness.
For anyone trying to understand drivers of purchase or loyalty, there are important lessons from the worlds of love and music about the differences between initial attraction and long-lasting relationships. Often it’s a difficult balancing act between the two. I have a much-loved Barbour International jacket that over the past 10 years is probably the best value item of clothing I’ve ever bought if I was to think rationally about £ per wear. However, whilst that has influenced me to purchase other items from the brand over the past decade, it undoubtedly wasn’t top of mind when I bought it in 2013.
There are though ways we can design our research to help us navigate such a complex area.
Firstly, what people tell us is important is often not the key area to focus on. In the world of quantitative research, techniques such as key driver analysis are relatively straightforward ways to avoid the pitfalls of stated importance questions but this discipline is equally important in the world of qualitative research too. With the increasing focus on speed, it’s all too tempting for clients to skimp on analysis time, especially if they’ve viewed the groups first-hand. Good qualitative research though brings far more to the table than reportage, shining a light on the needs and motivations behind what people say out loud.
Understanding the impact of when we run research is just as important. As our chart and relationship data shows us, a lasting ‘love at first sight’ is a relatively rare occurrence. For every 'Bohemian Rhapsody' in the record collection, there’s a 'Barbie Girl' (1.8 million copies sold in case you’re wondering) – proof that we don’t always make the right judgement calls in the heat of the moment. This is one of the reasons I’m hugely sceptical about the ever-increasing trend for putting a number to every single touchpoint – the scores we give immediately after a minor service touchpoint bear little relation to our behaviour longer term and therefore on the longer term commercial performance of a brand.
Finally, despite being sad enough to look up the peak chart positions of a list of songs, I’m the first to say that looking at data in isolation simply reiterates just how critical emotional factors are in any interpretation. We need to understand why songs which performed so poorly when first released have become classics of a generation. Why is it, for example, that 'Mr Brightside', a song which reached a relatively meagre No. 10 on release, has now passed 1 billion streams on Spotify? The answer won’t be found in data alone but by understanding it’s prominence in shared moments between friends - the guaranteed floor filler at the end of a wedding reception, for example. Equally, the role of films, TV shows (Glee, anyone?!) and even adverts in embedding a song in our consciousness shows just how important social currency is in both building and maintaining popularity.
As Tom Goodwin pointed out on Twitter last week, often to understand something we need to zoom out rather than zoom in. A quick look at the Peloton or WeWork stock price tells us that what can feel like a seismic shift at the time can feel like a blip on reflection. So next time you’re at risk of getting lost in that real time dashboard, take a step back and ask yourself if that datapoint you’re worrying about is more Mr Blobby than Mr Brightside….